### Author Topic: Matrix 9  (Read 1904 times)

0 Members and 1 Guest are viewing this topic.

#### ignatus

• Mature Member
• Posts: 481
• Thanked: 326 times
• Gender:
##### Matrix 9
« on: January 01, 2017, 04:14:03 AM »
This is a EC-bet, played with 2 ECs.

consider these two pairs/groups

RED HIGH<>BLACK LOW
RED LOW<>BLACK HIGH

(These pairs are also located on the same side of the wheel).

Trigger: When one of the pair within the group hit twice without the other pair being hit, bet the opposite pair with 2 ECs. (For an example RED HIGH has hit twice without BLACK LOW being hit, then bet BLACK & LOW EC).

Progression: double bets for each loss. This work pretty well, most i've seen is about 9 repeats within the same pair, so it Should cover most losses before reaching stoploss.

Bankroll: 100u playing with 1u bets, 500u playing with 5u bets.

4/4 Games won

As you can see in these charts, drawdowns didn't come even close to stoploss (-500u) at an extreme level of "bad luck" 7-8 Repeats of the same pair, this is an extreme event, although it didn't destroy our bankroll.

What is more, the reason why i like this bet is because it's played with 27 numbers, but in reality it's a 50/50 bet.

(Fist tests was played with 5u bets)

The following users thanked this post: december, Reyth

#### tickseeker

• New
• Posts: 37
• Thanked: 40 times
• Gender:
• It takes courage to be a Pig
##### Re: Matrix 9
« Reply #1 on: January 01, 2017, 10:14:15 AM »
I checked 100k  spins (real roulette spins) and found the following maximum streaks:
• be=11
• ro=10
• bo=8
• re=9
Here is the distribution

 BE RO BO RE 0 529544 528465 612332 570794 1 143946 143824 131861 138663 2 38566 38699 28483 33709 3 10500 10542 6104 8162 4 2855 2875 1411 1974 5 720 744 293 436 6 178 232 77 107 7 36 48 18 25 8 11 13 1 6 9 5 7 0 4 10 0 1 0 0 11 1 0 0 0 12 0 0 0 0 13 0 0 0 0 14 0 0 0 0

Edit
Haha. I did this for wrong EC pairs... well... should be the same
Edit2
100k spins
« Last Edit: January 01, 2017, 10:44:20 PM by tickseeker »

The following users thanked this post: ignatus, december, Reyth

#### ignatus

• Mature Member
• Posts: 481
• Thanked: 326 times
• Gender:
##### Re: Matrix 9
« Reply #2 on: January 01, 2017, 10:51:43 AM »
I checked 1 million spins (real roulette spins) and found the following maximum streaks:
• be=11
• ro=10
• bo=8
• re=9

Well, if that is true, calculate what trigger is needed (how many hits of one pair) I play with 2 hits (perhaps try 3 hits?) and a progression to cover the rest of these maximum steps, then you have your grail =)

Let's say you play with a trigger of 3 hits, then you need a 8 step martingale to cover the rest (3+8=11 steps total)

That would be 1-2-4-8-16-32-64-128 (8 steps)

OR perhaps your calculations are not correct(?) because it's a total hits in one pair, before the other hits (not how many times in a ROW a pair hits!)

Thanks!
« Last Edit: January 01, 2017, 11:23:16 AM by ignatus »

#### tickseeker

• New
• Posts: 37
• Thanked: 40 times
• Gender:
• It takes courage to be a Pig
##### Re: Matrix 9
« Reply #3 on: January 01, 2017, 11:27:40 AM »

The limit is shown in the graph's header. 100k *real* spins.

[BUG WARNING]

`_SECTION_BEGIN("ECer");/*  System Description:  - Follow BERO and BORE EC pairs  - When EC was hit 'n' times in a row then bet the pair's other side  - Progression: 2x after each miss until Hit AND then reduce back to 1*/// --- Initialize variablesLim = Param("Limit", 2, 2, 10);Li = Lim - 1;x = C - 100;r = C; r = 0;// Blacks in array is faster than "if-logic"blacks = C; blacks = 0;blacks[02] = 1; blacks[04] = 1; blacks[06] = 1; blacks[08] = 1;blacks[10] = 1; blacks[11] = 1; blacks[13] = 1; blacks[15] = 1;blacks[17] = 1; blacks[20] = 1; blacks[22] = 1; blacks[24] = 1;blacks[26] = 1; blacks[28] = 1; blacks[29] = 1; blacks[31] = 1;blacks[33] = 1; blacks[35] = 1; blacks[36] = 0;// --- Select data rangebi = BeginValue(BarIndex());ei = EndValue(BarIndex());if((NOT bi) OR (ei >= (BarCount-1))) {    bi = 1;  ei = 0;}// --- Subroutines// --- System Main Loopir = 0;max_be = max_ro = max_bo = max_re = -1;ctr_be = ctr_ro = ctr_bo = ctr_re = 0;pro_re = C; pro_be = pro_ro = pro_bo = pro_re = 0;for(i=bi; i<=ei; i++) {  be = blacks[x[i]] && !(x[i] & 1);  ro = !blacks[x[i]] && (x[i] & 1);  bo = blacks[x[i]] && (x[i] & 1);  re = !blacks[x[i]] && !(x[i] & 1);  // --- DO BETS  ir = 0;  if(ctr_be >= Lim) { if(!x[i]) ir -= (2^(ctr_be-Li))/2; else ir -= 2^(ctr_be-Li); if(ro) ir += 4^(ctr_be-Li); }  if(ctr_ro >= Lim) { if(!x[i]) ir -= (2^(ctr_ro-Li))/2; else ir -= 2^(ctr_ro-Li); if(be) ir += 4^(ctr_ro-Li); }  if(ctr_bo >= Lim) { if(!x[i]) ir -= (2^(ctr_bo-Li))/2; else ir -= 2^(ctr_bo-Li); if(re) ir += 4^(ctr_bo-Li); }  if(ctr_re >= Lim) { if(!x[i]) ir -= (2^(ctr_re-Li))/2; else ir -= 2^(ctr_re-Li); if(bo) ir += 4^(ctr_re-Li); }  // --- CALCULATE STATISTICS: Max streak lenghts and profile  // BE  if(be) {    ctr_be += 1;    if(ctr_be > max_be) max_be = ctr_be;  }  else {    pro_be[ctr_be] += 1;    ctr_be = 0;  }  // RO  if(ro) {    ctr_ro += 1;    if(ctr_ro > max_ro) max_ro = ctr_ro;  }  else {    pro_ro[ctr_ro] += 1;    ctr_ro = 0;  }  // BO  if(bo) {    ctr_bo += 1;    if(ctr_bo > max_bo) max_bo = ctr_bo;  }  else {    pro_bo[ctr_bo] += 1;    ctr_bo = 0;  }  // RE  if(re) {    ctr_re += 1;    if(ctr_re > max_re) max_re = ctr_re;  }  else {    pro_re[ctr_re] += 1;    ctr_re = 0;  }  r[i] = r[i-1] + ir;}// --- Plot resultsPlot(r, _DEFAULT_NAME(), colorWhite, styleDots);_SECTION_END();`
EDIT
Code bug (zero handling) fixed. Didn't update the graphs because the results didn't change very much (%).
Btw. sample size is 100k (not 1m).
« Last Edit: January 01, 2017, 01:26:53 PM by tickseeker »

The following users thanked this post: Reyth

#### ignatus

• Mature Member
• Posts: 481
• Thanked: 326 times
• Gender:
##### Re: Matrix 9
« Reply #4 on: January 01, 2017, 11:33:07 AM »
The topmost graphs looks amazing!

I got an error msg in RX, something about "system" in the beginning, if you could fix that?

Thanks alot! <3
« Last Edit: January 01, 2017, 11:40:33 AM by ignatus »

The following users thanked this post: tickseeker

#### tickseeker

• New
• Posts: 37
• Thanked: 40 times
• Gender:
• It takes courage to be a Pig
##### Re: Matrix 9
« Reply #5 on: January 01, 2017, 11:42:11 AM »
The topmost graphs looks amazing!

I got an error msg in RX, something about "system" in the beginning, if you could fix that?

Thanks alot! <3

Please check the Code before trying No Guarantee it's bug free

EDIT
Sorry I do not know RX. That code is for AmiBroker. I just did a check on other platform.

« Last Edit: January 01, 2017, 01:29:17 PM by tickseeker »

The following users thanked this post: ignatus, Reyth

#### ignatus

• Mature Member
• Posts: 481
• Thanked: 326 times
• Gender:
##### Re: Matrix 9
« Reply #6 on: January 01, 2017, 11:44:24 AM »
OK,
« Last Edit: January 01, 2017, 11:50:29 AM by ignatus »

#### ignatus

• Mature Member
• Posts: 481
• Thanked: 326 times
• Gender:
##### Re: Matrix 9
« Reply #7 on: January 01, 2017, 11:51:26 AM »
could you explain the style you're playing with the topmost graphs ?

#### tickseeker

• New
• Posts: 37
• Thanked: 40 times
• Gender:
• It takes courage to be a Pig
##### Re: Matrix 9
« Reply #8 on: January 01, 2017, 01:01:47 PM »
could you explain the style you're playing with the topmost graphs ?

Every graph is generated by the same code but using different streak length parameter. Rules can be read from the source code header (description)
« Last Edit: January 01, 2017, 01:27:40 PM by tickseeker »

#### tickseeker

• New
• Posts: 37
• Thanked: 40 times
• Gender:
• It takes courage to be a Pig
##### Re: Matrix 9
« Reply #9 on: January 01, 2017, 01:07:19 PM »
It seems there is a bug in zero handling.... Ill do the fix and update those graphs

The following users thanked this post: Reyth

#### MrPerfect.

• Hero Member
• Posts: 1815
• Thanked: 998 times
##### Re: Matrix 9
« Reply #10 on: January 01, 2017, 09:38:00 PM »
This is a EC-bet, played with 2 ECs.

consider these two pairs/groups

RED HIGH<>BLACK LOW
RED LOW<>BLACK HIGH

(These pairs are also located on the same side of the wheel).

Trigger: When one of the pair within the group hit twice without the other pair being hit, bet the opposite pair with 2 ECs. (For an example RED HIGH has hit twice without BLACK LOW being hit, then bet BLACK & LOW EC).

Progression: double bets for each loss. This work pretty well, most i've seen is about 9 repeats within the same pair, so it Should cover most losses before reaching stoploss.

Bankroll: 100u playing with 1u bets, 500u playing with 5u bets.

4/4 Games won

As you can see in these charts, drawdowns didn't come even close to stoploss (-500u) at an extreme level of "bad luck" 7-8 Repeats of the same pair, this is an extreme event, although it didn't destroy our bankroll.

What is more, the reason why i like this bet is because it's played with 27 numbers, but in reality it's a 50/50 bet.

(Fist tests was played with 5u bets)
Tendency looks nice. However these " critical events"  when bank go down are a bit scary. Probably something like laborshere progression to smooth it out? Or more refined trigger?

#### tickseeker

• New
• Posts: 37
• Thanked: 40 times
• Gender:
• It takes courage to be a Pig
##### Re: Matrix 9
« Reply #11 on: January 01, 2017, 10:52:20 PM »
On Limit 3 by risking 512 (max DD) units and getting 13572 (reward/risk 26.5) is pretty nice for the simple mechanical system I think.

The probkem here is 100k spins

Soo lets zoomin and start to tune rules... but im not gonna do that. Im buzy with other things.
« Last Edit: January 01, 2017, 11:09:48 PM by tickseeker »

The following users thanked this post: ignatus, Reyth